Difference in running time /distance / elevation data when exporting/importing different file formats from Garmin Connect (eg into Rubitrack)

I always export my runs from Garmin connect to finder(Mac) and then I import the file into Rubitrack (for Mac).

When importing the .fit file from Garmin connect I get the exact same running distance and running time as displayed on my Fenix 6X and on Garmin Express.

But when importing the .tcx or .gpx file I get a 100 meters more distance and 7 seconds time difference! My run shows as being 100 meters longer and 7 seconds longer (on a 5km run)

Any idea how this come?

NO

This is the abstract from an undergraduate paper I wrote some years back

Dr Phil said:

There was no significant difference between the distance (p=0.98) and time (p=0.99) at any of the three points of access.

For me it makes a difference if the distance displayed by the device during the run is 6.8 km and is also recorded in the FIT file. The real measured distance, however, is 7.1 km and a conversion of the FIT to GPX then also results in the real distance of 7.1 km. From this I must conclude that in this case you can not rely on the instantaneous distance traveled (Not because of bad GPS reception, but because of a "wrong" smoothing (which is even subsequently "repaired" by creating a GPX).The longer the distance traveled, the greater the negative deviation of the distance displayed on the watch, to the real distance traveled so far.

The point is that the FIT file is always shorter than the real distance in the GPX file. So it is not a problem of bad GPS reception, but a problem in the calculation of the data.
The question is why it is possible to get the correct distance by exporting the recorded data as GPX, but the FIT file recorded by the device during the run results in a too short distance.

Yes, the data is filtered, but why is it not possible in the FIT DIRECTLY to get the best possible data collection (data output) on distance ? Only the subsequent conversion to a GPX contains the correct track distance. The FIT does not correspond to the real distance is too short.
So a good file (GPX ) is created from a "bad" file (FIT) AFTER the fact.

Your study was done with six Garmin Forerunner 910XTs, two Garmin Forerunner 310XTs.
These devices probably still behaved differently than a current model like the fenix 6. I have read complaints from people who go back to the FR235 because these devices evaluate the current distance (and also pace) better.

If you connect the watch directly to the computer after a run, and evaluate the FIT file, it will contain a shorter (not real) distance than the GPX file generated later from it, which contains the real distance. Why ?

How is that possible? The FIT file already contains all captured data points.

Calculating the distance by the semicircles in the .FIT files (or by latitude and longitude in the .GPX files) can be done in different ways. The earth radius differs in different places on the globe and also the number of decimals when the calculations is done have impact. 3d distance or not etc. The number of decimals used for the latitude and longitude in the .GPX files is quite many. So, its not strange that the distance differs. I have also noticed this when doing my calculations regarding Speed/Pace.

The time differences maybe is a Total time vs Moving time thing.

AndersB said:

The time differences is probably a Total time vs Moving time thing.

This often cause confusion when activities are pushed (or uploaded to third party platforms). There are three times available in Garmin:

Time: The total time the timer was running (not paused intentionally or by Auto Pause feature)

Moving time: The total time of Time that you actually moved (auto calculated)

Elapsed time: The total time from when you started the activity until you stopped and saved. The during Resume Later will be included in this time

Different platforms choose to use different fields and Strava, as an example, choose depending on type of run. If it is a race it uses elapsed time, otherwise moving time.

Finally moving time seems to be a Garmin Connect calculation, not available in the FIT file. Probably post processing of the activity. Strava does the same and sometimes they are the same sometimes it differs (still within seconds), Strava has it's own way of deciding whether you were standing still or not.

https://support.strava.com/hc/en-us/articles/115001188684-Moving-Time-Speed-and-Pace-Calculations

Any idea how this come?

Yes. The reason is that the .FIT file contains the filtered/smoothed distance and enhanced_speed used for the graphs in GC. But when the .GPX or .TCX files are exported from GC, distance and enhanced_speed are not included.

Therefore, it's necessary that a third-party SW recalculate them from latitude/longitude for displaying trace or total distance, and the algorithms of the third-party SW differs. Some calculations are improved by topological maps.

An example:
When using a zigzag track from the GPS to calculate the distance, it is always longer than the direct straight distance between the start and end point. Therefore, the heading must be taken into account when calculating the distance.

Thank you for your answers, which are all plausible.
I am just very surprised that these facts are not more widespread.

Especially for marathon runners or trail runners, who cover longer distances, this should be an issue for them. If a marked course is exactly 100 km, but the watch at the finish line only shows 95 km ? Has no one wondered about this yet?

And there are reports that older models display the distance correctly. Why is it only with newer models that the distance is shortened ? And does this also affect other manufacturers? Does it have anything to do with the SONY chip ?

bluefish said:

Especially for marathon runners or trail runners, who cover longer distances, this should be an issue for them. If a marked course is exactly 100 km, but the watch at the finish line only shows 95 km ?

That is an error of 5% and I think that is near the limit of tolerable error. OTOH during a long run on a measured track you can't always run on the ideal line. So there is also some deviation from the course.

My first marathons with GPS units showed a longer distance, and it's very disappointing when the watch shows 42.195km and the finish line is still 2km ahead and you have exceeded the planned finish time :D

I think that the demand for accuracy has risen sharply in recent years and it is technically difficult to achieve. And may be, the low energy chip and antenna are not optimal for GPS signal reception. But if you run 100k or hike 3 days in a row, then you need a long battery life. So, you can't have everything at once.